Modification, Design, and Degradation Mechanisms of Nickel-Rich Layered Oxide Cathodes  for High-Capacity Lithium-Ion Batteries

Azhari, Luqman

Etd

Modification, Design, and Degradation Mechanisms of Nickel-Rich Layered Oxide Cathodes for High-Capacity Lithium-Ion Batteries

Public Deposited

To meet the energy capacity demands of next generation electric vehicles and other high energy use devices, significant developments in lithium-ion battery (LIB) cathodes will be required. Compositions which provide higher specific capacities need to be proven, and battery materials costs need to be reduced. Moving on from traditional cobalt-rich cathodes and towards nickel-rich compositions is the clearest and most impactful pathway towards these goals. As such, the implementation of transition metal layered oxide cathodes with high fractions of nickel is highly desired with capacities around 200mAh/g and has thus become a particularly popular topic of research in the past several years. By all expectations, commercial LIBs will be composed of nickel-rich cathodes within the next decade, provided that the inherent drawbacks and limitations are addressed. Herein, we choose to focus on LIB layered oxide cathodes with nickel contents of 80% and greater, and pathways towards practical industrial application. In Chapter 1, we establish the motivation and limitations of Ni-rich NMC cathodes and touch upon previously demonstrated designs towards resolving these limitations, specifically surrounding the typical polycrystalline secondary particle structure. In Chapter 2, we explore effects of extended aqueous processing of NMC cathodes of various compositions. Aqueous slurry processing is a desirable milestone for battery manufacturers, its success would eliminate the use of expensive and potentially harmful N-Methyl-2-pyrrolidone organic solvents. While isolating on the effects of water processing on solely the NMC cathode, longer processing timescales are utilized which are closer to those used in industry, contrary to shorter washing steps investigated in other works. We demonstrate that at these timescales, extensive fracture along grain boundaries occurs, with the severity of fracture positively correlated with the nickel fraction in the material. The removal of surface lithium residues and contact with water resulted in localized delithiation of the NMC surface and subsequent surface reconstruction along grain boundaries to rock-salt like ordering. It was also determined that the treated materials are more mechanically compromised, which adds risk for additional particle pulverization during calendaring processes. Lastly, electrochemical performance severely suffers from aqueous processing due to the resistive surface reconstruction layer and increased surface area, exhibiting higher capacity fade and reduced capacity at high charge/discharge rates. This work demonstrates the importance of surface modifications in order to utilize aqueous slurry processes at relevant scale. In Chapter 3, we look to develop a method to address the chemical instability and improve the performance of LiNi0.8Mn0.1Co0.1O2 (NMC811) while solely utilizing a coprecipitation method. NMC811 is well demonstrated to have reversible capacities ~200mAh/g when cycled to an upper voltage cutoff of 4.5V vs Li/Li+. However, surface-initiated instabilities caused by high valence Ni ions result in capacity fade, rendering it impractical for many applications. Based on the well documented electrochemical stability of Mn-rich oxides, we aimed for the synthesis of a heterogenous hydroxide precursor with a nickel-rich core and manganese-rich coating layer. Due to the importance of coprecipitation for high throughput synthesis, it was critical to maintain the entirety of the precursor synthesis within the coprecipitation reactor, with no post-precipitation modification steps outside of standard lithiation. This can be achieved through careful tuning of the reactor feedstocks and pH regulation, taking advantage of the high driving force for precipitation of transition metal hydroxides and the influence of pH on the coprecipitation process. After lithiation, the resulting cathode should be composed of an NMC811 core and Mn-rich spinel exterior. Commonly utilized electrochemical tests demonstrate a marked improvement in the electrochemical performance of the resulting cathode when compared to its uncoated version, especially in capacity retention at elevated temperatures and suppression of impedance growth. In Chapter 4, we explore the impact of minor amounts of lithium halides added to the lithium source during the calcination of NMC811. While extensive investigations have been conducted into cation doping of Ni-rich cathodes, less work is involved on anion doping. In this work, we demonstrate that minor additions of LiCl and LiBr result in significant morphology alterations, resulting in a porous cathode with ~3x more specific surface area than an undoped analogue. 5mol% Cl and Br doped samples exhibited significantly improved capacity retention and capacity under high charge/discharge rates, while F doping had negligible observed impact. The improvements in performance for Cl and Br doped samples were attributed to the onset of lithiation at lower temperature and formation of a compact and dense protective CEI layer upon cycling, which was investigated using surface sensitive techniques and supported by electron density-of-states calculations. This work demonstrates a new approach to Ni-rich cathode modifications through use of anion doping and tuning of the lithium source to affect the lithium melting temperatures. In Chapter 5, we attempt to synthesize large single-crystalline NMC811 cathode powders. NMC cathodes and other Ni-rich layered oxides are well documented to go through large anisotropic volume changes when brought to the highly charged state, with the structural changes occurring due to the significant level of withdrawal of lithium from the host structure during deep lithiation. Most NMC cathodes are typically comprised of micron sized spherical particles, which themselves are comprised of numerous nanosized primary particles with various orientations. Such changes in crystal lattice volume can initiate and exacerbate intergranular cracking in these polycrystalline secondary particles and lead to rapid capacity fade from electrical disconnection and continued exposure of reactive surfaces to electrolyte penetration. By removing the polycrystalline feature from the cathode, intergranular fracturing as a degradation mechanism is virtually eliminated. Single crystal NMC811 powders are synthesized through a facile molten salt method, utilizing lithium sulfate and sodium sulfate eutectics as a flux for crystal growth. The resulting powder is composed of micron sized single crystals of octahedral shapes. The shape and morphology are highly dependent on the initial molten salt content, composition, and sintering temperature/time. The eutectic salts are easily removed with water due to their high solubility, and the initial reversible capacity is comparable to typical polycrystalline NMC811 with excellent crystallinity. However, single crystal NMC811 suffers from rapid capacity fade, and failure mechanisms were investigated to be from intragranular cracking and planar gliding along the weaker (003) plane, which within the first cycle at high states of charge and exposes new surfaces to electrolyte-cathode interactions. Along with the growth of a thick CEI layer due to the more reactive (012) enclosed facets, cells suffer from rapidly increasing polarization at moderate levels of delithiation, with rapid irreversibility of the H2 to H3 transitions. This work demonstrates the effectiveness of flux growth for single crystal Ni-rich cathodes, while also exploring newly developed failure mechanisms and the need for further modifications for Ni-rich single crystal cathodes. In Chapter 6, we discuss recommendations and directions for future work, which includes preliminary data of doped single crystal cathodes and doped ultrahigh nickel-rich cathodes. In ultrahigh nickel-rich cathodes nickel content is pushed beyond 90mol% and cobalt content reduced to ~ 2.5mol%. In commercial considerations, increasing the Ni content and decreasing the Co content further lowers materials cost while improving reversible capacity. The resulting cathode of the composition LiNi0.95Mn0.025Co0.025O2 demonstrates a reversible capacity >200mAh/g when cycled to an upper voltage cutoff of 4.3V, which is higher than that of NMC811. Cyclic voltammograms demonstrate that increasing the nickel content to this extent results in much more severe transformations not observed in NMC811 or other cathodes with lower nickel content, with the H2 to H3 transition also occurring at a lower voltage. The much more reactive ultrahigh nickel-rich material thus exhibits severe capacity fade during galvanostatic charge/discharge cycling, with rapid fading attributed to a combination of mechanical vulnerability and surface-initiated degradation mechanisms leading to fast impedance growth. Usage of dopants is aimed towards improving the capacity retention of the cathode to acceptable level. Magnesium dopant at ~1mol% is demonstrated to be effective for suppressing capacity fade, especially when cycling at an elevated temperature of 60˚C. Additional improvements can be obtained through a combined doping and microstructural modification approach, with minor addition of boron being effective in tuning the primary particle growth. 1mol% of boron is effective in growing elongated, radially oriented primary particles from influence of boron atoms on the surface energy of the (003) plane and resulting in preferential orientations that are more effective in suppressing intergranular cracking

Creator